1 00:00:00,060 --> 00:00:01,200 2 00:00:07,440 --> 00:00:13,620 research project with the title Bittersweet 3 00:00:13,620 --> 00:00:19,020 brightness the impact of CEO's bright  personality traits on corporate fraud 4 00:00:22,020 --> 00:00:27,420 corporate fraud is an omnipresent topic that  often implies devastating outcomes for all 5 00:00:27,420 --> 00:00:32,820 societal members due to its tremendous  impact corporate fraud is a promising 6 00:00:32,820 --> 00:00:39,000 and exciting field of research let's look at two  companies that have committed fraud in the past 7 00:00:39,600 --> 00:00:46,500 or and Abercrombie and Fitch comparing both  companies by situational and organizational 8 00:00:46,500 --> 00:00:53,820 factors it becomes obvious that they were  completely different whereas Pfizer operated 9 00:00:53,820 --> 00:00:59,580 and of course still operates in the industry  of pharmaceuticals Abercrombie belongs to the 10 00:00:59,580 --> 00:01:05,460 apparel retail industry which is associated with  a completely contrasting industry environment 11 00:01:06,540 --> 00:01:12,600 moreover in the year before fraud commitment  both companies fundamentally differ in size 12 00:01:13,260 --> 00:01:18,780 Pfizer was about 22 times larger than  Abercrombie in terms of total revenues 13 00:01:20,460 --> 00:01:26,220 referring to their financial situation  abercrombie's return on assets was almost five 14 00:01:26,220 --> 00:01:33,120 times higher than pfizers comparing these cases  of corporate fraud situational and organizational 15 00:01:33,120 --> 00:01:40,200 factors show no commonalities that point to a  higher fraud propensity however the question 16 00:01:40,200 --> 00:01:46,920 arises are there commonalities on another  level looking at the respective CEO who was 17 00:01:46,920 --> 00:01:54,060 responsible for these fraudulent actions we ask  does the individual level matter more precisely 18 00:01:54,060 --> 00:02:01,020 do the CEOs of such firms have something in common  what makes them more prone to fraudulent actions 19 00:02:03,060 --> 00:02:07,320 so why do we regard the company's  CEOs as promising research Avenue 20 00:02:08,460 --> 00:02:14,280 upper echelon's theory states that CEOs are the  most powerful persons within their companies 21 00:02:14,940 --> 00:02:21,480 thus they have a high impact on organizational  decision making they Define a company's common 22 00:02:21,480 --> 00:02:26,700 vision and hence predetermined goals which  are building the base for strategic decisions 23 00:02:27,720 --> 00:02:33,240 furthermore like other individuals CEOs  differ in their perception of the environment 24 00:02:33,960 --> 00:02:40,620 these differences can arise from limited fields  of vision which means that different CEOs with 25 00:02:40,620 --> 00:02:45,900 different levels of experience and knowledge  attend differently to specific environmental 26 00:02:45,900 --> 00:02:51,840 stimuli more precisely they detect  different topics to put their focus on 27 00:02:52,680 --> 00:02:58,380 hence strategic decisions are based on  subjective interpretations of specific situations 28 00:03:00,240 --> 00:03:06,780 keeping these managerial differences between  CEOs in mind one crucial question arises 29 00:03:07,860 --> 00:03:10,200 where do these differences stem from 30 00:03:11,400 --> 00:03:18,480 upper echelon's Theory argues that managers  characteristics such as past experiences values 31 00:03:18,480 --> 00:03:25,500 and personality affect the firm's strategic  position and performance we're a CEO experience 32 00:03:25,500 --> 00:03:32,520 captures all relevant background information of a  CEO for example tenure or prior career experience 33 00:03:33,240 --> 00:03:39,960 CEO personality describes a relatively stable  disposition of character traits certain 34 00:03:39,960 --> 00:03:47,940 personality traits of a CEO strongly influence  her or his personal views of situations affect her 35 00:03:47,940 --> 00:03:54,660 or his decision making and consequently affect an  entire organizational behavior and its performance 36 00:03:56,940 --> 00:04:02,760 considering literature on corporate fraud  there are also three individual factors that 37 00:04:02,760 --> 00:04:09,900 strongly influence a CEO's Behavior first  a large pay gap between a firm's CEO and 38 00:04:09,900 --> 00:04:15,240 the other members of the top management team  can provoke a competitive culture and thus 39 00:04:15,240 --> 00:04:22,380 provide incentives for illegal means second a  CEO's willingness to engage in corporate fraud 40 00:04:22,380 --> 00:04:28,980 is also driven by the compensation structure  Scholars revealed a strong relationship between 41 00:04:28,980 --> 00:04:35,280 variable payments particularly equity-based  compensation and the likelihood of a lawsuit 42 00:04:36,600 --> 00:04:42,000 third addressing personality in corporate  fraud literature narcissism has been proven 43 00:04:42,000 --> 00:04:47,520 to foster a CEO's fraud propensity  narcissists intend to optimize their 44 00:04:47,520 --> 00:04:52,800 personal benefits and do not feel confused  by achieving this goal by acting unethically 45 00:04:54,900 --> 00:05:01,380 bringing together both streams of research it gets  obvious that personality traits might be promising 46 00:05:01,380 --> 00:05:07,980 in order to explain fraudulent Behavior up to  now a strong Focus has been on dark personality 47 00:05:07,980 --> 00:05:13,440 traits for example narcissism since dark  personality traits were strongly associated 48 00:05:13,440 --> 00:05:20,820 with detrimental company outcomes for example  fraud since Recent research has questioned the 49 00:05:20,820 --> 00:05:26,580 strict assignment of dark personality traits  to detrimental outcomes and bright personality 50 00:05:26,580 --> 00:05:34,080 traits to desirable outcomes we intend to break  with these predominant stereotypes therefore we 51 00:05:34,080 --> 00:05:40,860 connect bright personality traits with corporate  fraud in order to obtain a holistic view in 52 00:05:40,860 --> 00:05:48,600 this relationship we examine a complete set of  personality traits the Big Five personality traits 53 00:05:50,580 --> 00:05:56,460 based on an intensive literature screening we  developed the following five hypotheses referring 54 00:05:56,460 --> 00:06:03,000 to the impact of Big Five personality traits on  the CEO's willingness to engage in corporate fraud 55 00:06:04,200 --> 00:06:10,260 addressing the big five dimension open is to  experience we postulate that high levels of 56 00:06:10,260 --> 00:06:16,920 creativity unconventionality and an attraction  for risk-taking foster fraudulent activities 57 00:06:17,700 --> 00:06:23,760 hence the CEO's degree of openness to  experience is positively related to the 58 00:06:23,760 --> 00:06:31,320 likelihood to commit corporate fraud the second  dimension conscientiousness is associated with 59 00:06:31,320 --> 00:06:37,140 a strong goal in performance orientation which  seems to outweigh lower risk-taking preferences 60 00:06:38,040 --> 00:06:43,680 based on these findings we propose the  CEO's degree of conscientiousness to be 61 00:06:43,680 --> 00:06:51,000 positively related to the likelihood to commit  corporate fraud a strong ego and impulsive 62 00:06:51,000 --> 00:06:57,540 reaction to Temporary failure and the feeling  of invulnerability characterize extroverted CEOs 63 00:06:58,320 --> 00:07:05,280 these attributes might encourage extroverted CEOs  to engage in corporate fraud which is the reason 64 00:07:05,280 --> 00:07:11,640 why we hypothesize a positive relationship  between the CEO's degree of extraversion and 65 00:07:11,640 --> 00:07:17,340 corporate fraud for agreeableness  there arises an ambitious picture 66 00:07:17,940 --> 00:07:24,780 on the one hand agreeable CEOs prefer lower risk  taking which suggests a negative relationship 67 00:07:24,780 --> 00:07:30,960 between agreeableness and the engagement in  corporate fraud on the other hand agreeable 68 00:07:30,960 --> 00:07:36,540 CEOs could perceive High shareholder and  Broad pressure together with their avoidance 69 00:07:36,540 --> 00:07:42,060 of conflicts a positive relationship between  agreeableness and corporate fraud becomes obvious 70 00:07:43,140 --> 00:07:48,900 following this logic we introduce competing  hypothesis for the dimension agreeableness 71 00:07:51,420 --> 00:07:57,000 our last hypothesis refers to the last  dimension of the big five namely neuroticism 72 00:07:57,900 --> 00:08:03,900 neurotic CEOs show high levels of  self-doubt insecurity anxiety and hostility 73 00:08:04,800 --> 00:08:10,500 based on these attributes they show a strong  tendency for social comparison resulting in 74 00:08:10,500 --> 00:08:16,980 the perception of negative outcomes or threats  these perceived inabilities paired with the 75 00:08:16,980 --> 00:08:22,200 belief not to succeed by working hard could  push a neurotic CEO into fraudulent engagement 76 00:08:23,160 --> 00:08:28,620 consequently we hypothesize a positive  relationship between the CEO's degree 77 00:08:28,620 --> 00:08:33,000 of neuroticism and the corresponding  likelihood to commit corporate fraud 78 00:08:34,980 --> 00:08:40,500 in order to test our hypothesis we base our  investigation on a North American sample of 79 00:08:40,500 --> 00:08:50,400 186 CEOs of s p 500 firms our period  of Investigation is from 2002 to 2017. 80 00:08:51,300 --> 00:08:55,260 but how did we collect data for  the CEO's personality assessment 81 00:08:56,640 --> 00:09:02,400 as already mentioned we chose the personality  concept of the Big Five in order to investigate 82 00:09:02,400 --> 00:09:06,960 personality traits holistically and  break with predominant stereotypes 83 00:09:08,160 --> 00:09:13,560 Oscars and it all complains that CEO's decisions  based on their behavioral mechanisms are 84 00:09:13,560 --> 00:09:19,500 investigated insufficiently due to methodological  obstacles in assessing their personality traits 85 00:09:20,340 --> 00:09:25,740 since primary data is predominantly only  available for small and private companies 86 00:09:25,740 --> 00:09:33,120 these findings could not be easily transferred to  large public ones an alternative to primary data 87 00:09:33,120 --> 00:09:40,620 could be found in language samples of CEOs current  approaches refer to spoken language in form of the 88 00:09:40,620 --> 00:09:49,080 CEO's answers given in conference calls conference  calls consist of two parts a presentation as well 89 00:09:49,080 --> 00:09:55,740 as a questions and answers part in line with prior  research we intended to gather the corresponding 90 00:09:55,740 --> 00:10:01,740 transcripts of quarterly earnings conference  calls for the CEOs by using Thomson Reuters icon 91 00:10:03,060 --> 00:10:07,440 since the presentation part of those earnings  conference calls is often accused of being 92 00:10:07,440 --> 00:10:15,360 scripted we based our investigation on the answers  part of the Q a session in our last step we had 93 00:10:15,360 --> 00:10:21,120 to choose the methodological approach in this  study we leveraged approaches from the field of 94 00:10:21,120 --> 00:10:28,260 machine learning to measure a CEO's personality  in line with my Umbra at all we trained a neural 95 00:10:28,260 --> 00:10:32,880 network based on the stream of consciousness  essay data set provided by panabaga and King 96 00:10:34,140 --> 00:10:40,980 we choose an open vocabulary and test the  different engram sizes from 1 to 5 as features for 97 00:10:40,980 --> 00:10:47,940 the model we used compact model architecture with  three hidden layers and five neurons in each layer 98 00:10:48,780 --> 00:10:54,300 the train model was then used to predict  the individual personality dimension of a 99 00:10:54,300 --> 00:11:02,280 CEO for all big five dimensions in our study we  conceptualized our dependent variable as the CEO's 100 00:11:02,280 --> 00:11:07,920 engagement in corporate fraud operationalized  a set of Securities class actions lawsuits 101 00:11:08,880 --> 00:11:13,200 Securities class actions lawsuits  are shareholder initiated lawsuits 102 00:11:13,200 --> 00:11:18,420 against firms and their top Executives in  response to misleading or deceptive Behavior 103 00:11:19,560 --> 00:11:26,100 as top executives are directly responsible for  disclosing information to shareholders this 104 00:11:26,100 --> 00:11:29,640 type of fraud is attributable  to top executive defendants 105 00:11:30,840 --> 00:11:37,140 in line with previous studies corporate fraud  is binary coded one if a CEO was engaged in 106 00:11:37,140 --> 00:11:45,000 any settled lawsuit in a given year and zero  otherwise we collected cases of Securities 107 00:11:45,000 --> 00:11:50,220 class actions lawsuits at the Securities class  action Clearinghouse of Stanford Law School 108 00:11:51,900 --> 00:11:59,340 the conviction of a CEO's illegal behavior is  compared to all non-fraudulent CEOs a relatively 109 00:11:59,340 --> 00:12:05,520 rare case thus random sampling of fraudulent  and non-fraudulent CEOs is not reasonable 110 00:12:06,720 --> 00:12:11,280 following prior research we used a more  powerful matched pair sample design 111 00:12:12,360 --> 00:12:19,860 firms corresponding to the CEOs were matched based  on three criteria fraud year industry and firm 112 00:12:19,860 --> 00:12:26,400 size we matched each firm accused of fraud with  a control firm from the same industry based on 113 00:12:26,400 --> 00:12:32,640 the four-digit standard industry classification  code firm size was represented by total assets 114 00:12:35,460 --> 00:12:39,900 we used a conditional logic regression  analysis to test our hypothesis 115 00:12:40,560 --> 00:12:46,320 this method is typically used in matched pair  sample designs with binary dependent variables 116 00:12:46,320 --> 00:12:51,960 as it considers the conditional nature of the  likelihoods that a matched pair sample represents 117 00:12:52,800 --> 00:12:57,420 conditional logistic regressions do  not produce an overall intercept and 118 00:12:57,420 --> 00:12:59,520 thus estimate a model with a fixed effect 119 00:13:01,560 --> 00:13:06,120 after the wrap up regarding our research  model let's have a look into our results 120 00:13:07,020 --> 00:13:14,520 model 1 solely includes the control variables of  our study in this setting performance relative to 121 00:13:14,520 --> 00:13:21,060 the social aspiration level has a negative effect  on the incident of corporate fraud this is in line 122 00:13:21,060 --> 00:13:25,920 with Harrison bromley's results that illustrate  that the likelihood of fraudulent behavior 123 00:13:25,920 --> 00:13:32,280 is highest for the lowest or most negative  values of negative social relative performance 124 00:13:34,260 --> 00:13:40,440 model 2 includes our independent variables  the big five dimensions of personality traits 125 00:13:41,220 --> 00:13:47,340 considering the output table it gets obvious that  the higher the level of a CEO's conscientiousness 126 00:13:47,340 --> 00:13:54,600 the higher is her or his propensity to perpetrate  fraud the corresponding results of our estimation 127 00:13:54,600 --> 00:14:01,860 show a positive and significant relationship  referring to a CEO's degree of neuroticism 128 00:14:01,860 --> 00:14:07,200 also a positive relationship to the likelihood  to commit corporate fraud could be detected 129 00:14:08,040 --> 00:14:13,800 the estimated coefficient for this independent  variable is positive and statistically significant 130 00:14:13,800 --> 00:14:21,540 providing good support for our hypothesis 5.  consequently CEOs who score high on neuroticism 131 00:14:21,540 --> 00:14:27,540 have a higher predicted probability to engage  in behavior that results in unsettled lawsuit 132 00:14:28,620 --> 00:14:34,380 for the other three dimensions openness to  experience extroversion and agreeableness 133 00:14:34,380 --> 00:14:39,960 we found no significant results which led to  the rejection of the corresponding hypothesis 134 00:14:41,820 --> 00:14:46,260 so which contributions to our  study offer to existing literature 135 00:14:46,980 --> 00:14:53,220 first we provide further insights into upper  echelon's Theory by focusing on the CEO while 136 00:14:53,220 --> 00:14:59,400 investigating her or his impact on strategic  decision making more precisely corporate fraud 137 00:15:00,600 --> 00:15:06,960 previous research on fraudulent Behavior often put  the firm including its structures and culture in 138 00:15:06,960 --> 00:15:14,280 the focus of their investigations however in most  of the cases the top Executives of a company are 139 00:15:14,280 --> 00:15:20,820 accused to perpetrate fraud with our study we  follow the call of schnatterly at all to shed 140 00:15:20,820 --> 00:15:27,120 more light on the CEO by examining the impact  of her or his personality on corporate fraud 141 00:15:27,960 --> 00:15:32,940 second we contribute to the ongoing discussion  about the impact of bright and dark personality 142 00:15:32,940 --> 00:15:39,420 traits on organizational outcomes whereas previous  Studies have shown positive as well as negative 143 00:15:39,420 --> 00:15:46,380 organizational consequences associated with dark  personality traits Smith at all complains about 144 00:15:46,380 --> 00:15:52,800 prevailing literature to strongly assign bright  personality traits solely to beneficial outcomes 145 00:15:53,580 --> 00:15:58,440 in order to question this relationship  and break with traditional stereotypes 146 00:15:58,440 --> 00:16:03,180 we counter-intuitively connect bright  personality traits measured by the big 147 00:16:03,180 --> 00:16:11,580 five Dimensions with detrimental company outcomes  more precisely fraud by doing so we derive an 148 00:16:11,580 --> 00:16:16,320 in-depth understanding about the composition  and mechanisms of the single elements of the 149 00:16:16,320 --> 00:16:21,720 Big Five personality traits and illustrate  their impact on managerial decision making 150 00:16:23,100 --> 00:16:29,160 third linguistic analysis is widely used  for personality assessment purposes since 151 00:16:29,160 --> 00:16:34,680 it represents a remarkable alternative to  questionnaires and interviews to the best 152 00:16:34,680 --> 00:16:40,440 of our knowledge no other study related this  measurement approach which facilitates broader 153 00:16:40,440 --> 00:16:47,460 personality samples to corporate fraud by  pursuing an open vocabulary approach followed 154 00:16:47,460 --> 00:16:54,120 by feed forward neural networks as well as boosted  regression trees we refer an Innovative and modern 155 00:16:54,120 --> 00:17:01,620 methodological approach for assessing personality  to the field of fraudulent Behavior fourth our 156 00:17:01,620 --> 00:17:07,200 study contributes to fraud literature since  we analyze a specific kind of fraud Securities 157 00:17:07,200 --> 00:17:13,260 class actions lawsuits that attracted not only  practitioners but also scholars in recent years 158 00:17:14,040 --> 00:17:19,560 with our study we follow the emerging trend of  applying these lawsuits in management literature 159 00:17:20,340 --> 00:17:24,960 since Securities class actions lawsuits  are raised against a company and its top 160 00:17:24,960 --> 00:17:29,940 Executives they are perfectly suitable  for research on top management teams 161 00:17:31,500 --> 00:17:37,140 besides theoretical contributions we also  support companies and Regulators with new 162 00:17:37,140 --> 00:17:42,600 intriguing insights about the degree of certain  personality traits and its impact on the tendency 163 00:17:42,600 --> 00:17:49,680 to behave fraudulently considering the example of  the introduction both of the depicted fraudulent 164 00:17:49,680 --> 00:17:56,280 CEOs Henry mckinnell and Michael Jeffries show  very similar levels for all big five dimensions 165 00:17:57,300 --> 00:18:01,080 the degree of agreeableness and  neuroticism is even identical 166 00:18:02,340 --> 00:18:08,640 based on these insights companies are enabled to  search for or choose more appropriate candidates 167 00:18:08,640 --> 00:18:15,240 for vacancy or positions organizations need  to maintain The credibility in their corporate 168 00:18:15,240 --> 00:18:21,360 governance systems by improving internal controls  and developing an ethical corporate culture 169 00:18:22,200 --> 00:18:28,860 the CEO of a company should represent this  culture particularly with her or his personality 170 00:18:30,540 --> 00:18:35,160 simultaneously our results support  controlling institutions like the 171 00:18:35,160 --> 00:18:39,780 Securities and Exchange Commission  sec in the context of fraud prevention 172 00:18:40,560 --> 00:18:46,080 legal Regulators are highly interested in  antecedents of corporate fraud to concentrate 173 00:18:46,080 --> 00:18:50,700 their investigations on cases that show  high indications for illegal schemes 174 00:18:52,560 --> 00:18:58,500 with our study we hope to encourage Scholars as  well as practitioners to investigate corporate 175 00:18:58,500 --> 00:19:06,300 fraud and especially its antecedents in order to  reduce or in the best case to stop its occurrence 176 00:19:08,220 --> 00:19:13,140 if you have any questions or remarks we  would be delighted to receive your message